|Home | About | Journals | Submit | Contact Us | Français|
Toxoplasma gondii is an obligate intracellular parasite capable of causing fatal infections in immunocompromised individuals and neonates. Examination of the phosphatidylserine (PtdSer) metabolism of T. gondii reveals that the parasite secretes a soluble form of PtdSer decarboxylase (TgPSD1), which preferentially decarboxylates liposomal PtdSer with an apparent Km of 67 μm. The specific enzyme activity increases by 3-fold during the replication of T. gondii, and soluble phosphatidylserine decarboxylase (PSD) accounts for ~20% of the total PSD, prior to the parasite egress from the host cells. Extracellular T. gondii secreted ~20% of its total PSD activity at 37 °C, and the intracellular Ca2+ chelator 1,2-bis(2-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid tetrakis (acetoxymethyl ester) inhibited the process by 50%. Cycloheximide, brefeldin A, ionic composition of the medium, and exogenous PtdSer did not modulate the enzyme secretion, which suggests a constitutive discharge of a presynthesized pool of PSD in axenic T. gondii. TgPSD1 consists of 968 amino acids with a 26-amino acid hydrophobic peptide at the N terminus and no predicted membrane domains. Parasites overexpressing TgPSD1-HA secreted 10-fold more activity compared with the parental strain. Exposure of apoptotic Jurkat cells to transgenic parasites demonstrated interfacial catalysis by secreted TgPSD1 that reduced host cell surface exposure of PtdSer. Immunolocalization experiments revealed that TgPSD1 resides in the dense granules of T. gondii and is also found in the parasitophorous vacuole of replicating parasites. Together, these findings demonstrate novel features of the parasite enzyme because a secreted, soluble, and interfacially active form of PSD has not been previously described for any organism.
Toxoplasma gondii is an obligate intracellular parasite of the phylum Apicomplexa, which also includes the human and animal parasites Plasmodium, Eimeria, Neospora, and Babesia (1, 2). Toxoplasma causes an opportunistic disease, toxoplasmosis, in individuals with immune dysfunction and in developing fetuses and neonates. The tachyzoite stage of T. gondii infects host cells and replicates to high numbers, ultimately lysing the cell, prior to the next round of invasion. Successful intracellular infection by T. gondii tachyzoites depends on its ability to modify the external micro-milieu by secreting a variety of factors (3, 4). Toxoplasma harbors at least three distinct secretory organelles, which contribute to the parasite invasion and formation and modification of an intracellular parasitophorous vacuole (PV).3 Micronemes and rhoptries, located at the apical end of T. gondii, discharge their contents at the time of invasion and formation of the nascent PV (3, 4). The dense granules are released into the vacuolar space after invasion is complete. A coordinated secretion of these organelles is a prerequisite for successful invasion and replication of T. gondii within its host cell.
Phosphatidylserine decarboxylase catalyzes the synthesis of phosphatidylethanolamine (PtdEtn) from phosphatidylserine (PtdSer), a crucial process required for membrane biogenesis by numerous prokaryotic and eukaryotic organisms (5, 6). The PSD enzymes belong to an interesting subgroup of decarboxylases that contain a pyruvoyl prosthetic group constituting an essential part of the catalytic site. The mature enzyme is a heterodimer composed of a large membrane-associated β-subunit and a small pyruvoyl-containing α-subunit (6, 7). The mature form is generated by auto-proteolytic cleavage of a single polypeptide producing α- and β-subunits and covalent attachment of the pyruvoyl group in a concerted reaction. The bacterial enzyme is an integral protein present in the cytoplasmic membrane of Escherichia coli (7). Two distinct PSDs are expressed in Saccharomyces cerevisiae; one is located in inner mitochondrial membrane (ScPsd1p) and the other in Golgi and vacuolar compartments (ScPsd2p) (6, 8). In mammals, only one PSD associated with the mitochondrial membrane has been described (8). Only two parasite PSD enzymes have been studied in any detail so far (9, 10).
Our previous studies demonstrated the major routes of phospholipid synthesis in T. gondii (11). The parasite can acquire the precursors, serine, ethanolamine, and choline, from its environment and use them for the synthesis of its major phospholipids, PtdSer, PtdEtn, and phosphatidylcholine, respectively (11). These precursor utilization studies in conjunction with enzyme assays identified the presence of the Kennedy pathways for synthesis of PtdEtn and phosphatidylcholine (8, 12, 13), the base-exchange pathway for generation of PtdSer (8, 14), and a PtdSer decarboxylase route for formation of additional PtdEtn (7, 8, 15) in T. gondii (11). Surprisingly, the PSD activity (214 nmol/h/mg of protein) was 10-fold higher than that observed in extracts prepared from other eukaryotes such as yeast (16) and mammalian cells (17). This report focuses on TgPSD1, one of two detectable PSD activities, expressed in T. gondii tachyzoites. TgPSD1 is secreted to the vacuole of intracellular parasites and into the external environment of host-free (axenic) parasites. TgPSD1 secreted by extracellular parasites is a soluble protein, which acts on liposomal and host cell PtdSer.
Dulbecco's modified Eagle's medium (DMEM), MEM amino acids, and vitamins were purchased from Invitrogen. The l-[U-14C]serine and l-[1-14C]serine were obtained from ICN Radiochemicals Inc. Brefeldin A, colchicines, cytochalasin D, and cycloheximide were obtained from Sigma. The intracellular calcium chelator BAPTA-AM was purchased from Molecular Probes. Lipids were procured from Avanti Lipids. Silica Gel H and Sil60 plates for thin layer chromatography (TLC) were from Analtech and Merck. The E. coli strain XL-1 Blue (Stratagene) was used for molecular cloning and vector amplification. Parasite RNA isolation, cDNA synthesis, and plasmid preparations were performed with Invitrogen kits. All DNA-modifying enzymes and primers were obtained from New England Biolabs and Invitrogen. Primary (anti-HA) and secondary (Alexa488 and Alexa594) antibodies were from Invitrogen. The TaTi strain of T. gondii was a kind donation of Dominique Soldati-Favre (University of Geneva, Switzerland). Transgenic T. gondii tachyzoites expressing red fluorescent protein and the pNTP3 expression vector were kindly provided by Isabelle Coppens (The Johns Hopkins Bloomberg School of Public Health, Baltimore). Anti-TgGra1, anti-TgGra3, and anti-TgGra5 antibodies were gifts of Marie-France Cesbron-Delauw (CNRS, Grenoble, France). Jurkat cells were kindly provided by Carsten G. Lüder (Georg-August-University, Göttingen, Germany).
Human foreskin fibroblasts (HFF) obtained from the American Type Culture Collection were cultured in DMEM supplemented with 10% fetal calf serum (FCS), 2 mm glutamine, MEM nonessential amino acids, 100 units/ml penicillin, and 100 μg/ml streptomycin at 37 °C in a humidified incubator with a 10% CO2 atmosphere (11). The HFF were passaged by trypsinization at least once a week and used up to the 10th passage. T. gondii tachyzoites of the RH strain were routinely propagated in vitro by serial passage in HFF monolayers at a multiplicity of infection (m.o.i.) of 4, unless stated otherwise. To purify the axenic parasites, supernatants from freshly lysed monolayers were collected and centrifuged at 2000 × g for 10 min followed by three subsequent washings with cold phosphate-buffered saline (PBS). For some preparations, infected but unlysed HFF monolayers were passed twice through 20- and 22-gauge needles to release tachyzoites. The parasites were used immediately after isolation. Although T. gondii tachyzoites are unable to replicate outside of host cells, they remain viable long enough (t½ ~10 h) in an extracellular environment to permit experimentation. In the early stages of these studies, we compared parasites purified on Nycodenz gradients (18) with those prepared by simple centrifugation and PBS washing. We observed no significant differences between the two methods and hence used the simple centrifugation and washing procedure for the experimental manipulations described in this report. Parasite homogenates were prepared from a suspension containing 1–2 × 108 tachyzoites/ml by probe sonication at 0 °C using five 30-s bursts at 50 watts with 30-s cooling intervals between bursts. The homogenates were kept on ice prior to initiating enzyme reactions by pre-mixing all assay components at 0 °C and then shifting to 37 °C. Heat-inactivated (95 °C for 10 min) enzyme extract and/or reactions lacking enzyme were included as negative controls for catalysis in each assay.
PSD activity was measured by trapping 14CO2 released from Ptd[U-14C]Ser or Ptd[1′-14C]Ser on filter paper impregnated with 2 m KOH (7). Dioleoyl-Ptd[U-14C]Ser and Ptd[1′-14C]Ser were synthesized from l-[U-14C]serine or l-[1-14C]serine and dioleoyl-CDP-diacylglycerol using PtdSer synthase. The PtdSer synthase was purified from E. coli strain JA-200 harboring the plasmid pPS3155 as described previously (19). The reactions were performed in 16 × 100-mm borosilicate glass tubes sealed with an air-tight rubber septum, to which was attached a well holding the 2 m KOH-saturated paper. The parasite extract was prepared in 50 mm potassium phosphate buffer (pH 6.8), 0.25 m sucrose, 0.5 mm phenylmethylsulfonyl fluoride (PMSF) and 3 mm EDTA. The 0.8-ml assay mixture contained 60 mm potassium phosphate (pH 6.8), 0.17 m sucrose, 0.35 mm PMSF, 2 mm EDTA, 0.5 mm 2-mercaptoethanol, 0.5 mm dioleoyl-Ptd[U-14C]Ser (0.1 μCi/μmol), 0.1% (w/v) Triton X-100, and 0.2 ml of parasite extract. In studies using liposomal forms of Ptd[U-14C]Ser, the detergent was omitted. The liposomes were freshly prepared using a LiposoFast (Avestin) and 100-nm filters. The enzyme reactions were terminated at the indicated times by the addition of 0.5 ml of 0.25 m H2SO4, introduced through the rubber septum using a hypodermic needle. The emitted 14CO2 was trapped for 30 min prior to recovering the filter paper for liquid scintillation counting.
To measure the PSD expression during intracellular parasite growth, the HFF were infected at a multiplicity of 3. The infected cells were washed with cold PBS and harvested by scraping at 8, 16, 24, 32, 40, and 48 h post-infection. The crude extract and reactions were prepared as described above. The 0.5-ml reactions contained 50 mm potassium phosphate (pH 6.8), 0.125 m sucrose, 0.25 mm PMSF, 1.5 mm EDTA, 0.5 mm 2-mercaptoethanol, 0.25–0.5 mm (0.4–0.8 μCi/μmol) of dioleoyl-Ptd[U-14C]Ser, and 0.2 ml of parasite homogenate and were conducted for 1 h at 37 °C.
Aliquots of 1 × 108 parasites were incubated with 1–100 μm (0.1 mCi/reaction) dioleoyl-Ptd[U-14C]Ser in 1 ml of intracellular-type medium (ICM) in 16 × 100-mm borosilicate glass tubes at 37 °C for 1–6 h with shaking. The ICM contained 20 mm HEPES, 140 mm KCl, 10 mm NaCl, 2.5 mm MgCl2, 5 mm glucose, 0.1 μm CaCl2, 1 mm sodium pyruvate, MEM vitamin solution lacking choline, MEM amino acids, and serine-free nonessential amino acids (pH 7.4). The nonessential amino acid mix contained 200 μg/ml of Ala, Asp, Glu, Gly, Pro, and Asn. The lipid precursors serine, choline, ethanolamine, and inositol (20 μm each) and ATP-Mg (1 mm) were also included.
Typically, 1–1.3 × 108 parasites were incubated at 37 °C for 0.25–4 h in 1.3 ml of ICM supplemented with 20 mm each of serine, choline, ethanolamine, and inositol and 0.75 mm freshly prepared ATP-Mg. ATP-depletion experiments were performed in glucose- and pyruvate-free ICM, lacking ATP and supplemented with 7.5 mm NaN3, 7.5 mm NaF, 384 mm NaVO4, 153 mm dinitrophenol, and 7.5 mm oligomycin A. The incubation conditions for 14CO2 trapping and measurements were the same as those described above. A low speed supernatant was prepared by three centrifugations (1500 × g for 15 min). Subsequently, a high speed supernatant (HSS) was prepared by centrifuging the low speed supernatant at 150,000 × g for 45 min. An aliquot of 0.9 ml of HSS was mixed with 0.1 ml of 1% Triton X-100-solubilized or liposomal Ptd[U-14C]Ser. The enzyme reaction was incubated for 1–6 h at 37 °C and then terminated by the addition of 0.5 ml of 0.25 m H2SO4. The 14CO2 was trapped for 30–60 min in the sealed reaction tubes prior to recovering the filter paper for liquid scintillation spectrometry.
Lipids were extracted using the Bligh and Dyer method (20). Each 1-ml PSD reaction was terminated by addition of 1.1 ml of CH3OH and 1.1 ml of CHCl3, followed by vigorous mixing and centrifugation. The resultant chloroform phase was washed three times with 2.1 ml of CH3OH/PBS/CHCl3 (10:9:1.5, v/v). The final chloroform phase containing lipids was dried, and the radioactivity was quantified by liquid scintillation spectrometry. Alternatively, lipids were dried under N2 and suspended in 50–100 μl of CHCl3/CH3OH (9:1, v/v) for TLC analysis. Lipids were analyzed by one-dimensional TLC on Silica Gel H plates in CHCl3, CH3OH, 2-propanol, KCl (0.25%), triethylamine (90:28:75:18:54, v/v) or by two-dimensional TLC on Sil60 plates (first dimension in CHCl3/CH3OH/NH4OH (65:35:5, v/v) and the second dimension in CHCl3/CH3COOH/CH3OH/H2O (75:25:5:2.2, v/v)). Lipids were visualized by spraying TLC plates with 0.2% (w/v) anilino-1-naphthalenesulfonic acid and exposure to UV light, iodine staining, or autoradiography. All lipids were identified based on their co-migration with authentic standards.
HFF were infected with T. gondii tachyzoites at an m.o.i. of 3 and cultured for 48 h. Parasites in the early stage of lysis were harvested by scraping cells and then liberated from host cells by two passages through 22- and 27-gauge needles. The parasites were washed by three centrifugations in diethylpyrocarbamate-treated PBS. RNA was isolated using the TRIzol method (Invitrogen) and transcribed into first-strand cDNA. The TgPSD1 cDNA containing a C-terminal HA epitope was amplified using the forward (CTCGATATCATGGCTAAGGTTATGAGGCTTATC) and reverse (CTCTTAATTAATCAAGCGTAATCTGGAACATCGTATGGGTAGAGATCCCCATTGGTAAGCA) primers and Pfu-Ultra FusionII polymerase (Stratagene). The cDNA was cloned into the pNTP3 vector using EcoRV and PacI restriction enzyme sites. Purified tachyzoites (1 × 107) of the TaTi strain (21) were transfected with 50 μg of the TgPSD1-HA construct using the BTX630 instrument (2 kV, 50 ohm, 25 microfarads, 250 μs). The drug-resistant parasites were selected with 1 μm pyrimethamine as described before (22), and clonal transgenic tachyzoites were analyzed for PSD expression.
The indirect immunofluorescence analysis of intracellular and extracellular T. gondii tachyzoites was performed as described previously (23). In brief, the parasitized HFF monolayers grown on glass coverslips for 24–36 h were washed with PBS and fixed with 4% paraformaldehyde for 10 min, followed by neutralization in 0.1 m glycine/PBS (5 min). Cells were permeabilized in 0.2% Triton X-100/PBS for 20 min, and nonspecific binding was blocked with 2% BSA in 0.2% Triton X-100/PBS. Samples were stained with primary antibodies (rabbit anti-HA 1:1500; mouse anti-HA 1:1000; mouse anti-TgGra1 1:500; rabbit anti-TgGra3 1:500; mouse anti-TgGra5 1:500), followed by three washes with 0.2% Triton X-100 in PBS. The corresponding secondary antibodies (mouse or rabbit Alexa488 or Alexa594) were applied (1:3000) for 45 min, and after three PBS washes, the slides were mounted in DAPI-Fluoromount G for fluorescent imaging (Apotome, Carl Zeiss, Germany). For immunostaining of the extracellular stage, the parasites were allowed to settle on poly-l-lysine-coated coverslips before fixation by paraformaldehyde (4%) for 10 min at room temperature. Samples were stained and imaged as described above.
Jurkat cells (E6.1 T-lymphocyte derived cells) were grown in RPMI medium supplemented with 10% FCS and penicillin (100 units/ml)/streptomycin (100 μg/ml) at 37 °C and 5% CO2 in a humidified incubator. For induction of apoptosis, 1 × 106 cells were resuspended in 1 ml of fresh RPMI medium containing 0.1% DMSO either with or without 1 μm staurosporine (Invitrogen) for 2 h in a 12-well plate. Cells were harvested by centrifugation (400 × g for 10 min at room temperature) and washed twice with fresh medium. Subsequently, the cells were incubated with wild type or a transgenic T. gondii strain overexpressing TgPSD1-HA (m.o.i. ~30) or with the medium control for 2 h. The cells were resuspended in 100 μl of annexin V-FLUOS labeling reagent (Roche Applied Science), stained for 15 min in the dark, and then fixed with 2% paraformaldehyde for 15 min. Cells were subsequently washed and resuspended in 200 μl of HEPES buffer for FACS analyses or seeded on poly-l-lysine-coated coverslips for microscopy. Samples were examined for annexin V staining on an LSRFortessa cell analyzer (BD Biosciences). The parasites and Jurkat cells were fractionated prior to analyzing the results using FlowJo suite.
As an intracellular parasite, T. gondii imports numerous classes of molecules to support its growth. To investigate whether T. gondii can transport and metabolize exogenous phospholipid, we incubated the axenic parasites with unilamellar liposomes of dioleoyl-PtdSer and measured its decarboxylation. Surprisingly, parasites exhibited a high rate of concentration- and time-dependent decarboxylation (Fig. 1, A and B). The presence of high PSD activity in the living parasite is consistent with previous studies conducted with the parasite homogenates (11). Initially, these assays suggested import of PtdSer by the parasite and decarboxylation in T. gondii. However, further analyses did not support this interpretation of the data.
We next sought to determine where liposomal PtdSer was being decarboxylated. For these experiments, a defined number of parasites were incubated in ICM followed by preparation of cell-free supernatants. Surprisingly, we observed a time-dependent increase in PSD activity in the HSS indicating the presence of an enzyme that appeared to be secreted by the parasite (Fig. 2A). The secretion was rapid in the 1st h and gradually reached a plateau over the next 3 h. The unexpected presence of PSD in the HSS after parasite incubation demonstrated the soluble nature of the enzyme. Our assays using extracellular-type (high Na+/K+) medium yielded no differences in secretion of PSD compared with ICM (data not shown). To exclude the prospect of parasite lysis during culture, we performed similar experiments with transgenic T. gondii expressing red fluorescent protein and measured its release into the medium. We did not find any measurable release of red fluorescent protein that might be attributed to cellular disintegration. Moreover, the number of intact parasites prior to and after incubation was the same. The observed behavior of the enzyme is consistent with this secreted PSD being a soluble protein.
To further establish the nature of the parasite-secreted enzyme as PSD, we examined the product of Ptd[U-14C]Ser metabolism by tachyzoite-derived HSS. As shown in Fig. 2B, the major product of PSD was Ptd[U-14C]Etn as deduced by thin layer chromatography and radioactive imaging of the phospholipid products. As expected, we also observed an intense band of PtdSer, which represents the substrate excess. Our simultaneous measurement of PtdSer decarboxylation rates, deduced by the release of 14CO2, and of Ptd[U-14C]Etn formation detected by TLC, matched the theoretical yields of the two products in a ratio of ~1:2, suggesting that no other significant product was formed during the HSS-mediated catalysis (Fig. 2C). These data confirm that the catalytic function of this novel PSD is not different from other reported PSD proteins.
Based on our conjecture that a soluble PSD may act at membrane interfaces, we compared the catalytic behavior of PSD using detergent-solubilized and liposomal PtdSer (Fig. 3A). The secreted PSD of T. gondii consistently exhibited nearly 2-fold higher activity with liposomal PtdSer compared with a detergent-dispersed substrate. However, the total PSD activity in the parasite homogenates was 1.5-fold higher with the detergent-dispersed substrate compared with the liposomal PtdSer. The activity of the secreted soluble PSD with liposomal PtdSer was linear for 2 h (Fig. 3B). The secreted PSD decarboxylated PtdSer in a concentration-dependent manner, and about 5% of the substrate was consumed at 200 μm PtdSer. Kinetic analysis indicated an apparent Km value of 67 μm with liposomal PtdSer. Taken together, these results indicate a relatively stable secreted T. gondii PSD, with an increased affinity for liposomal PtdSer. In addition, these data suggest the presence of a second membrane-bound PSD in T. gondii, which is consistent with the parasite database (ToxoDB) predicting two PSD genes.
To determine some of the general characteristics of PSD secretion, we examined the temperature, ATP, and calcium dependence of the process, as shown in Fig. 4. Under routine conditions at 37 °C, the extracellular parasites secreted about ~20% of their total PSD activity (Fig. 4A). The enzyme secretion was inhibited by almost 90% when the incubation temperature was reduced to 0 °C. As anticipated, the secretory process was ATP-dependent, and we observed an ~92% reduction in T. gondii-secreted soluble PSD activity, when the parasites were preincubated with metabolic inhibitors depleting the intracellular ATP pool (Fig. 4B). When the parasites were exposed to the intracellular Ca2+ chelator BAPTA-AM (Fig. 4C), the secretion of PSD was inhibited by ~50%. We were unable to reverse the effect of BAPTA-AM by simultaneous addition of calcium (1 mm). The addition of the calcium ionophore, ionomycin, did not affect the enzyme secretion by free parasites (data not shown). Together, the data show a partial dependence of the PSD secretion on intracellular calcium stores in free parasites. The use of ATP inhibitors and BAPTA-AM did not have any direct effect on the PSD activity (data not shown). The dependence of PSD secretion on temperature, ATP, and intracellular calcium is consistent with enzyme release from intracellular stores.
We examined the expression of PSD during the intracellular growth of the parasite in human fibroblasts and discovered an increase in specific activity of PSD as the parasite number increases. As shown in Fig. 5A, the total PSD activity in cultures of infected fibroblasts increases in a linear manner from 8 h after infection until 32 h, prior to the host-cell lysis and parasite egress. Between infection and parasite-induced lysis of the host cells, we observed a 4-fold increase in the specific activity of PSD. In contrast, the PSD activity in uninfected host cells was very low and remained constant between 8 and 32 h. We also measured the specific activity of the soluble and membrane-associated PSD pools in HSS and high speed pellets in parasites mechanically liberated from host cells at 16 and 30 h after infection. The results in Fig. 5B demonstrate the PSD activity was distributed in the pellet as well as in the soluble fraction, and the latter included ~20% of the total enzyme activity at 16 and 30 h of infection. Just prior to egression (~30 h), the soluble PSD-specific activity was 3-fold higher than that at 16 h post-infection. Our data strongly suggest the presence of two different variants of PSD in the parasite, one is membrane-associated and the other is soluble. This latter result also suggests that the soluble pool of PSD is likely secreted from the parasite during and/or after egress from the host.
To determine whether there is a soluble PSD pool stored in host cell-free axenic T. gondii tachyzoites, we conducted secretion assays in the presence of cycloheximide, an inhibitor of protein synthesis in T. gondii (24). The cycloheximide treatment failed to inhibit PSD secretion. We also examined the effect of brefeldin A, an inhibitor of Golgi-mediated secretory function, upon release of PSD. This drug treatment failed to block PSD secretion, consistent with the idea that the parasite stores significant quantities of the enzyme in prepackaged secretory organelles. Treatment of axenic parasites with colchicine or cytochalasin D also did not affect PSD secretion, indicating that the process is likely not directed by microtubule or microfilament function.
Inspection of the T. gondii database revealed two putative PSD genes. Our experimental annotation of TgPSD1 identified a protein containing 968 amino acids with a likely signal peptide at the N terminus, as shown in Fig. 6A. The SignalP suite predicted a cleavage site between positions Val-26 and Gln-27 (score, 0.742) of TgPSD1, consistent with the observed secretory behavior. A PtdSer decarboxylase homology domain encompassing residues 419–720 is conserved in the TgPSD1. Analysis of the secondary structure of TgPSD1, devoid of the signal peptide, did not reveal any hydrophobic transmembrane peptides. The catalytic site of TgPSD1 is predicted within a 681FGST684 motif, which conforms to a consensus subunit cleavage site reported for other PSD enzymes (5, 6). Autocatalytic processing of the TgPSD1 proenzyme (~108 kDa) should yield a pyruvoyl-containing α-subunit (~32 kDa) derived from the C terminus and a β-subunit (~76 kDa) derived from the N terminus.
Bioinformatic analyses of TgPSD1 revealed two putative N-glycosylation sites at positions Asn-94 and Asn-820. Numerous sites for O-glycosylation, C-mannosylation, and phosphorylation are also predicted, but their significance remains to be determined. Phylogenetic analyses showed that TgPSD1 might be distantly related to type I PSD enzymes from mammals, fungi, and plants, whereas type II and bacterial PSD proteins form their own clade as depicted schematically in Fig. 6B. Our attempts to express an active form of TgPSD1 in E. coli and S. cerevisiae were unsuccessful. However, TgPSD1 could be overexpressed in its active form in T. gondii tachyzoites as described below.
To examine the subcellular location of the TgPSD1 protein in parasites, we expressed this protein containing a C-terminal HA epitope (TgPSD1-HA) in T. gondii tachyzoites under the control of the pTgNTP3 promoter (Fig. 7). Immunostaining of the transgenic parasites within the host cell revealed that TgPSD1-HA was secreted into the parasitophorous vacuole, which separates the entire population of replicating parasites from the host cytosol. Fig. 7 demonstrates that the secreted TgPSD1-HA co-localized with TgGra1, TgGra3, and TgGra5 (Fig. 7A), which are bona fide proteins of the parasite dense granules that are secreted into the vacuole (25). TgGra1 remains soluble within the vacuolar space, whereas TgGra3 and TgGra5 are associated with the vacuolar membranes. When the focal plane was selected to demonstrate the distribution of TgGra1, there is significant co-localization of this protein with TgPSD1-HA. When the focal plane was chosen to accentuate the localization of TgGra3 and TgGra5, TgPSD1-HA was also detectable within the dense granules and at the vacuolar membrane. Fluorescence images of parasites after egress from host cells (Fig. 7B) revealed co-localization of TgPSD1-HA with TgGra1 and TgGra3. These results confirm that TgPSD1-HA resides in the dense granules of T. gondii, prior to its secretion into the vacuole by replicating parasites. Consistent with our secretion studies, these findings also show that dense granule pools of TgPSD1-HA are present in axenic parasites.
We next examined the function of TgPSD1 by determining the catalytic activity and secretory nature of the recombinant protein. The TgPSD1-HA was overexpressed in transgenic parasites. Immunoblot analysis confirmed the presence of TgPSD1-HA in parasite homogenates, HSS fractions, and in high speed pellet (HSP) fractions containing membranes and dense granules. As shown in Fig. 8A, TgPSD1-HA was detectable almost exclusively as the processed, mature form of the ~32-kDa α-subunit, indicating that nearly all of the proenzyme was efficiently processed to the mature enzyme.
We quantified the overexpression of TgPSD1 by measuring the enzyme activity in subcellular fractions isolated from parasitized HFF (Fig. 8B). The transgenic parasites overexpressing TgPSD1-HA had ~10-fold more PSD activity in homogenates compared with the parental strain, and ~20% of the enzymatic activity was detectable in a soluble form present in HSS. Secretion studies were also performed using the axenic transgenic parasites. As shown in Fig. 8C, the transgenic strain of T. gondii secreted ~10 times more soluble TgPSD1 than the parental strain, and neither strain released a significant amount of TgPSD1-HA at low temperature. Collectively, the data confirm the activity of TgPSD1-HA and the soluble and secreted properties of the enzyme.
In an effort to understand the biological role of TgPSD1, we examined the ability of the secreted enzyme to act upon PtdSer exposed on cell surfaces. In these experiments, we compared the amounts of externalized PtdSer present on mammalian cells treated with staurosporine, in either the absence or presence of axenic parasites secreting TgPSD1. Staurosporine elicits an apoptotic host cell response that induces PtdSer externalization at the plasma membrane (26). The exposed PtdSer was detected by measuring fluorescent annexin binding (27). Cultures of DMSO-treated control Jurkat cells contained a minor population of cells that stained with fluorescent annexin V, detected by microscopy and cell sorting (Fig. 9, A and B). Treatment of Jurkat cells with staurosporine elicited fluorescent annexin staining in nearly the whole population (Fig. 9A). When the drug-treated cells were exposed to wild type parasites for a period of 2 h, the annexin binding of the cells remained unchanged (parental versus medium in Fig. 9B). In contrast, a co-incubation with transgenic T. gondii overexpressing TgPSD1 caused a significant 34% reduction in surface intensity of staurosporine-treated cells. These experiments reveal that under the appropriate conditions the TgPSD1 can reduce the amount of surface-exposed PtdSer present on host cells. From a biochemical perspective, the data show that the interfacial catalysis performed by TgPSD1 with a liposomal substrate is recapitulated with PtdSer present in the exofacial leaflet of plasma membranes.
This study provides evidence for the presence of a novel phosphatidylserine decarboxylase expressed by T. gondii that is soluble and secreted from the parasite into the PV during the intracellular phase of parasite growth. During the axenic phase of the parasite life cycle, the TgPSD1 is secreted into the external medium. Although some soluble forms of PSD have been reported in bacteria (28), and as a subpopulation of a Plasmodium transgene encoded protein expressed in yeast (10), no secreted forms of the enzyme have been reported previously from any organism. Our results also suggest a second membrane-associated PSD is present in the parasite.
Unlike homologs from other organisms residing in either the inner mitochondrial membrane or in Golgi/vacuolar/endosomal membranes (5, 6), TgPSD1 harbors a putative secretory signal peptide with a predicted proteolytic cleavage site at its N terminus but no predicted transmembrane domains. The secretion of PSD by free T. gondii tachyzoites is time-, temperature-, and ATP-dependent, and it appears to be independent of the exogenous PtdSer concentration, Na+ or K+ gradients across the plasma membrane, active protein synthesis, trafficking from endoplasmic reticulum to Golgi network, and cytoskeletal integrity. These features are common to proteins secreted from parasite dense granules (25, 29). Immunolocalization of TgPSD1 in the dense granules and in the PV of parasites further supports the secretory nature of TgPSD1 in axenic parasites. Because dense granule release is not directional, as in other T. gondii secretory organelles, such as micronemes and rhoptries, there is no specific requirement for a cytoskeleton-dependent transport of vesicles to the apical end of T. gondii for PSD secretion.
The dense granules in T. gondii constitute an unusual secretory pathway that allows soluble export of many membrane-associated proteins (e.g. TgGra proteins) to the PV via information encoded within the N-terminal domain (25, 29). These proteins insert into the PV membranes after initially being secreted as soluble proteins into the vacuole. We show that intracellular parasites release TgPSD1 into the PV, where the enzyme may bind to membranes to support the biogenesis of an expanding vacuole and ensure a faithful parasite replication. The product of TgPSD1 catalysis, PtdEtn, can promote protein conformational changes and influence lateral movement and activity of membrane-bound proteins (30, 31). The association of TgPSD1 with the PVM also has the potential to regulate sequestration of host-derived endocytic vesicles (32) or provide an optimal bilayer environment for vacuolar proteins. Finally, TgPSD1-mediated PtdEtn production can also influence membrane curvature that may dictate the fusion and fission events in the PV. To this end, the catalytic action of TgPSD1 could facilitate escape of the parasite by altering the integrity of the vacuolar and host membranes.
Unlike many membrane-bound enzymes of lipid metabolism, TgPSD1 does not require detergents to interact with its substrate, and it shows nearly twice the activity with liposomal PtdSer when compared with detergent-dispersed substrate. These findings are consistent properties for an enzyme that works at a membrane interface. Our studies with transgenic parasites and apoptotic cells show that the overexpressed and secreted TgPSD1 can reduce the levels of PtdSer exposed on apoptotic cells. Although this same result was not obtained with parasites expressing wild type levels of TgPSD1, this finding may be due to the differences between the catalysis by TgPSD1 in a relatively large cell-free volume in tissue culture, as compared with the concentrated foci of infected cells in solid tissues. Typically, 64–128 parasites are released from a parasitized host cell, and the egressed population of parasites acts very locally in time and space to infect the adjacent cells. In such a microenvironment, the wild type levels of TgPSD1 are expected to be sufficient for reducing the amount of externalized PtdSer on host cells.
Externalized plasma membrane PtdSer is a potent signal that stimulates phagocytic cells to recognize, engulf, and destroy host cells with this surface property (33). In vivo, the catalytic action of TgPSD1 may serve to suppress this PtdSer signal on host cells, generated either in response to parasite invasion or following neighboring host cell lysis. A reduction of surface PtdSer should function to reduce or even prevent the cells from detection and phagocytosis by macrophages and thereby promote persistence of the parasite. In brief, the extracellular TgPSD1 may enhance the ability of T. gondii to evade detection by the innate immune system.
In summary, our studies identify an unusual PSD family member expressed by T. gondii that is soluble and secreted and is capable of catalysis at membrane interfaces. TgPSD1 can reduce the exposed PtdSer content on apoptotic cells. These findings identify unexpected new aspects of the phospholipid metabolism of T. gondii.
We thank Matthew R. Hepworth (Humboldt University, Berlin, Germany) for assistance with the measurements and analyses of the FACS data.
*This work was supported, in whole or in part, by National Institutes of Health Grants AI030060 and 2R37GM32453 (to D. R. V.). This work was also supported by German Research Foundation Grants SFB618 and GRK1121 (to N. G. and R. L.).
The nucleotide sequence(s) reported in this paper has been submitted to the GenBankTM/EBI Data Bank with accession number(s) JN003619.
3The abbreviations used are: